LEARN. PRACTICE. GROW.1. Distinguish effort, capacity, and duration
In a role split between production and projects, someone may work eight hours a day while having only two available for an initiative. Remaining effort does not shrink because the calendar shows more days. For divisible work, constant productivity, and no waits, divide effort by available daily capacity. Twelve hours at three hours per day require four working days. This calculation is a model, not a promise: interruptions, environment access, dependencies, and uninterrupted-work requirements may increase duration. Record assumptions and confirm availability with whoever manages the resource. A generic allocation percentage is less useful than a calendar showing periods that are actually available.
2. Read dependencies before adding
Consider A taking two days, followed by B taking five and C taking three. B and C can run in parallel. D requires both complete and takes two days. With sufficient resources and finish-to-start relationships without additional lags, total duration is two plus the longer branch plus two: nine days. Path A-B-D drives completion in this model. Adding everything gives twelve days and ignores parallel work. Selecting only the longest activity gives five and ignores dependencies. For real calendars, include nonworking days, justified constraints, and availability; a simplified example’s result must not be presented as a guaranteed production date.
3. Resolve the resource conflict
Now change one condition: B and C need the same full-time specialist and cannot overlap. Logic permits parallel work, but the resource prevents it. With no other resources or breaks, B and C occupy eight sequential days, making the total twelve. Either order produces that total in this example, although it may change risks and intermediate milestones in a real project. Discuss alternatives with explicit impact: another qualified person, technically divisible work, a scope change, or a revised deadline. Do not shorten durations just to make a chart fit a date. The forecast should represent an executable sequence, including constraints that the logical diagram alone does not show.
4. Keep commitment and forecast visible
The baseline records the approved commitment; the forecast represents the best current reading of the data. If the baseline finishes on day 20 and the updated plan points to day 24, present both, the cause, and response options. An approved change may justify revising the baseline through the applicable process. Until then, hiding day 24 reduces transparency; silently replacing day 20 removes the reference. Where a risk response has an owner and an observable trigger, monitor that trigger. If a supplier fails to confirm access by the agreed deadline, activate the planned alternative and check execution rather than waiting until delay becomes inevitable.
5. Adapt an iteration without hiding quality
In a Scrum context, the Sprint Goal gives coherence to selected work. Developers adjust the plan as they learn and collaborate with the Product Owner when scope needs renegotiation. If more testing work emerges, consider removing an item less necessary to the goal while retaining the Definition of Done. Do not use plan changes as a reason to declare incomplete work finished. The Daily Scrum helps plan near-term work around impediments; the Sprint Review supports discussion of outcomes and new needs with stakeholders. These Scrum-specific mechanisms do not define every hybrid approach. The project still needs decisions on budget, external dependencies, and transition.
6. Complete transition and measure the outcome
Migrating an application does not automatically demonstrate that support can operate it. If transition requires recovery performed by the RUN team, record that criterion, its owner, and evidence. A supported session with synthetic data can reveal uncertainties that emailing a manual will not expose. Also distinguish delivery from benefit: automating reconciliation creates a capability; reducing time under comparable conditions demonstrates an outcome. If the first value improves while volume is half its usual level, present it as an early signal and plan suitable observations. In the final exercise, combine the revised schedule, pending criteria, and measurement plan in a short update for the sponsor.
Example: A=2, B=5, C=3, D=2. A precedes B/C; D waits for both. Sufficient resources: 9 days. One exclusive specialist for B/C: 12 days. The change concerns executable capacity, not historical effort.
Common pitfalls
Adding parallel activities, ignoring a shared resource, editing the baseline to erase variance, lowering quality to preserve a list, and confusing installation with operational autonomy.
Related topics: Decompose, sequence, and estimate · Learn and adapt delivery
A useful forecast combines logic and capacity. Adapt work transparently, preserve quality criteria, and measure benefits under conditions that permit comparison.
Reference: Schedule Assessment Guide · CAPM ECO 2023; official outline consulted 2026-09-29